The Reaction between Quinones and Metallic Enolates. XIII

taine,19 but not the creatine,16 methyls may be .... Trimethylethylquinone and Sodium Malonic Ester. 933 group .... melted at 250°, and its ethyl est...
4 downloads 0 Views 545KB Size
LEE IRVIN SMITHAND J. W. OPIE

932

Of particular interest are the experiments of du Vigneaud who has reported that in young rats the methionine methyl may be transferred to choline and to creatine6 and that the cholinea and betaine,lS but not the creatine,l6 methyls may be transferred t o homocystine. Acknowledgments.-We wish t o acknowledge financial assistance from the Theelin Fund administered b y the Committee on Grants for Research of St. Louis University. We are also indebted to Merck and Co., Inc., for a supply of choline chloride and to Parke, Davis and Co., Inc., for a supply of Natola. (19) Chandler and du Vigneaud, J . Biol. Chen., 186,223 (1940).

Vol. 63

Summary 1. Hemorrhagic degeneration is the result of a dietary deficiency of choline and of the labile methyl supply. 2. Betaine, like methionine, contributes to the labile methyl supply of the body and may be substituted for choline. The effectiveness of betaine corresponds to the utilization of only one of the three methyl groups. 3. Creatine does not contribute to but does spare the labile methyl supply. 4. Cysteine, homocystine and glutathione, but not taurine, increase the severity of hemorrhagic degeneration. ST. LOUIS,MISSOURI

RECEIVED NOVEMBER 1, 1940

[CONTRIBUTION FROM THE SCHOOL OF CHEMISTRY OF THE UNIVERSITY OF MINNESOTA]

The Reaction between Quinones and Metallic Enolates. XIII. Trimethylethylquinone and Sodium Malonic Ester' BY LEE IRVW SMITHAND J. W. OPIE~

Trimethylbromoquinone (I) reacts with sodium malonic ester to produce 3-carbethoxy-6-hydroxy5,7-dimethyl-8-bromocoumarin(II).3 This coumarin is the only product of the reaction; the bromine atom is unaffected by the enolate, but it does exert a strong directing influence upon the formation of the pentad-enol system (111) from 0

K

O

0

I, R=Br V, R=CzH5

11, R = B r

111, R = B r IV I, since of the three methyl groups, the one meta t o the bromine atom reacts exclusively. I n the previous paper' it was shown that while no product resulting from attack of sodium malonic ester (1) Paper X I I , Taxa JOURNAL, 68, 612 (1941). (2) Abstracted from a thesis by Joseph W. Opie, presented t o the Graduate Faculty of the University of Minnesota, in partial fulfillment of the requirements for the Ph.D. degree, July, 1940. (3) Smith and Johnson, THIS JOURNAL, SO, 673 (1937).

upon the bromine atoms of dibromo-m-xyloquinone could be isolated, the reaction leading t o the coumarin (IV) was difficult to control and, moreover, the coumarin I V differed from all its analogs previously studied in the great ease with which the heterocyclic ring could be opened. In dibromo-m-xyloquinone, there is no methyl group meta to either of the bromine atoms, and i t appeared, when the results obtained with this quinone were compared those obtained with the monobromoquinone I, that the nature as well as the orientation of the substituents in the quinones exert a pronounced effect not only upon the course of the reaction, but also upon the ease with which it occurs and upon the stability of the heterocyclic ring in the coumarin formed. In order to explore these effects further, a study has been made of the reaction between trimethylethylquinone (V) and sodium malonic ester. Having shown that tetraethylquinone was inert toward the enolate, i t was assumed that the ethyl group of V would not react, and that the reaction with sodium malonic ester would be confined to the methyl groups. Each of the three methyl groups thus became a potential point of entry for the malonic ester residue, and three isomeric coumarins could result, depending upon the direction and magnitude of the orienting effect of the ethyl

April, 1941

'TRIMETHYLETHYLQUINONE AND SODIUM

MALONIC ESTER

933

coumarin ester gave no solid acetate, but resulted in an oil which could not be crystallized. The broad range in the melting points of both the ester and the acid, and the fact that the ester CHa 0 gave excellent analytical values in spite of its poor melting point, indicated that the substance was a HHo\ 3 C r f\/ i = 0COOCzHj CHo\/\/ zHif)~&~CzHI mixture of a t least two coumarins. When the CzH6 CH, ester (m. p. 148-165') was dissolved in benzene VI VI1 and chromatographed on calcium carbonate, a C2H6 0 narrow dark brown zone formed a t the top of the tube, but most of the material remained unadsorbed. The benzene filtrate, on evaporation, gave material which melted a t 172-179'. Sub\TI11 stitution of aluminum oxide for calcium carbonSince, moreover, the ethyl group is not as highly ate led to a definite yellow layer in the lower part of polar as a bromine atom, i t was considered quite the tube, and from this zone an ester melting a t probable that reaction of the quinone V would not 181-182' was obtained. As the ester was fairly insoluble in ether, experibe confined exclusively t o any one of the methyl ments were performed t o determine whether or not groups, as it was in the case of quinone I, but that the mixture could be separated by fractional solua mixture of any two, or even all three of the coution in this solvent. No difference was found bemarins VI, VI1 and VI11 might result. tween the dissolved and undissolved portions of When sodium malonic ester was added t o the the substance, but i t was found that repeated quinone V in benzeneJ4there was formed, in excrystallization of the substance from ether-pecellent yield, an insoluble red sodium derivative. troleum ether ultimately gave a product which After removal of this sodium derivative, the filmelted a t 183-184', while very slow recrystallizatrate, when cooled, deposited the hydroquinone tion from alcohol ultimately gave a product meltcorresponding to V. When the filtrate from the ing a t 185-188O. The mother liquors from etherred salt was steam distilled, up to 40% of the starting material V could be recovered from the petroleum ether solutions, on standing in a redistillate. These results showed that the qui- frigerator for two months, deposited material This material was none V, like duroquinone, reacted in such a manner which melted a t 153-166'. washed several times with ether and the ether w a that half of the quinone was reduced to hydroevaporated. The residue, after crystallization quinone and that one mole of the quinone theofrom petroleum ether, melted a t 150-152' and had retically should give half a mole of coumarin and the composition ClaH1sOr. half a mole of hydroquinone. Upon this basis, Hydrolysis of the 152' ester gave a n acid CI4the yield of red sodium derivative was practically HMOS which melted a t 230-232', and acetylation quantitative, a result quite in contrast with that of the ester gave an acetate CISHZ~OS which melted shown by dibromo-m-xyloquinone. a t 151-153', but which, when mixed with the The red sodium derivative, when suspended in 152' ester, melted a t 125-130'. Hydrolysis of ethanol and subjected to the action of hydrothe 185' ester gave an acid which melted a t 237chloric acid, gave a yellow coumarin in over 90% yield. This coumarin, after crystallization from 238', but acetylation of this ester produced an ethanol, melted a t 148-165' and gave analytical acetate having the wide melting point range of values agreeing closely with the formula ClaHls05 141-160' and it was not possible to obtain a (VI, VI1 or VIII). Hydrolysis of this coumarin purer product by recrystallization. Thus chromatographic adsorption, combined ester with hydrochloric acid5 led to a coumarin with partial solution and crystallization. led t o acid which melted at 192-210'. Neither crystaltwo coumarin esters with fairly narrow melting lization of this acid from acetic acid, nor precipipoint ranges: A, melting a t 185-188', constituttation of i t from its solution in sodium carbonate, ing about a third of the crude coumarin ester, and changed its melting point. Acetylation of the B, melting a t 150-152', constituting a much (4) Smith and Dobrovolny, TRISJOURNAL, 48, 1093 (1926). smaller part of the crude coumarin ester. Mix( 5 ) Smith and Denves. ibid.. 68. -. 304 - - - f\ -l Q - -3 - ,6.) .

group. Thus coumarins VI, VI1 and VIII, respectively, would be formed b y reaction a t the methyl groups numbered 1 , 2 and 3 in V.

934

LEE IRVIN SMITHAND J. W. OPIE

Vol. 63

tures of A and B melted a t intermediate tempera- is an inseparable mixture of two of the coumarin tures. The wide difference in melting points be- esters (A = VI VII) while the other (B) is an tween A and B appeared to be unusual, but not inseparable mixture of all three coumarin esters impossible, for isomers so closely related as are It follows, therefore, that the ethyl group in this VI, VI1 and VIII. Each of the esters, on hy- quinone does not possess in this reaction, the drolysis, gave acids with fairly sharp melting strong directing effect shown by the bromine from B, 230-232'), points (from A, 237-238'; atom in trimethylbromoquinone I. but only B gave an acetate which had a narrow That the esters VI, VI1 and VIII, differing only melting point range (151-153'). in the position of very similar alkyl groups in posiThe separation of the crude coumarin ester tions 5 , 7 and 8, should prove to be inseparable isinto the components A and B was extremely te- while disappointing-not particularly surprising dious and time consuming, and the amounts of and it has been observed before in cases in which a the two components obtained were not sufficient reaction produces isomers so closely related as are for degradative studies. Consequently, in order the esters VI, VI1 and VIII.7 to settle the question as t o the nature of A and B, Experimental Part* the three coumarin esters VI, VI1 and VI11 were synthesized.6 3-Carboxy-6-hydroxy-ðyl-7,8- 3,6-Dinitro-1,2,4,5-tetraethylbenzene.-The hydrocardimethylcoumarin melted a t 223-224' and its bon (3 g.)g was vigorously shaken with sulfuric acid (75 cc.) until solution was complete. A mixture of fuming ethyl ester (VI) melted a t 178-179.5'. 3-Car- nitric acid (20 cc., d. 1.5) and sulfuric acid (15 cc.) was boxy - 6-hydroxy- 7 -ethyl - 5,8 - dimethyl coumarin added in portions and the mixture was vigorously shaken. melted a t 250', and its ethyl ester (VII) melted After the addition was complete, the solution was warmed a t 199-201 '. 3-Carboxy-6-hydroxy-8-ethyl-5,7-(70-80") for a few moments, during which oxides of dimethylcoumarin melted at 232-234', and its nitrogen were evolved. The solid nitro compound separated during the reaction. The mixture was poured over ethyl ester (VIII) melted a t 173-174.5'. ice (300 g.) and the solid was removed and crystallized from The method of thermal analysis, using the pure ethanol. The product formed long white needles (2.15 g., compounds, shows that neither A nor B is a pure 56%) which melted a t 149-151'. Smith and Harris,ID substance. Further, the data indicate that B is a who prepared this compound by nitration of hexaethylmixture of all three esters, for the melting points benzene, report it t o melt a t 143-145'. 3,6-Diamino-1,2,4,5-tetraethylbenzene Stannichloride. of mixtures of B and each of the three pure esters -The above dinitro compound (3.2 9.) was dissolved in lie above 152', suggesting that as the concentra- boiling acetic acid (60 cc.) in a large Aask. The flame tion of any one of the pure components is increased, was removed, and a hot solution of stannous chloride dithe melting point of the mixture with B tends t o hydrate (20 g.) in hydrochloric acid (100 cc.) was added shift toward that of the added component. The all a t once. After the initial very vigorous reaction subsided, the mixture was refluxed for thirty minutes and fact that a mixture of the three pure esters melts then was cooled in an ice-salt-bath. The stannichloride a t 153-160', that is, a t about the same tempera- was filtered off and washed successively with cold hydroture as any of the binary mixtures containing B, chloric acid, water, alcohol and ether. The product further indicates that B is composed of all three of (7.78 g., 98Q/,) could be crystallized from water. Tetraethyl-p-benzoquinone.-The above stannichloride the esters. A, on the other hand, appears to be a (7.8 8.) was added to a solution of ferric chloride (20 g.) in mixture composed only of VI and VII, for all of water (75 cc.). After standing for twenty-four hours, the binary mixtures of these three substances melt the mixture was steam distilled and the distillate was exa t 180-185', while there is a great depression tracted with ether. Removal of the ether left 2.24 g. produced in the melting point of A, VI or VI1 (go%, based upon the nitro compound) of quinone which a t 60-62O." when any of these substances is mixed with B or melted Tetraethylquinone and Enolates of Malonic Ester.-A. with V I I I i . e., there is no VI11 in A. Ethyl malonate (1 cc.) was added to powdered sodium Hence the reaction between trimethylethyl- (0.15 g.) suspended in benzene (50 cc.) and the mixture quinone (V) and sodium malonic ester leads to a was heated until the metal had reacted completely. The mixture containing all three of the coumarin es- quinone (0.5 8.) in benzene (15 cc.) was added and the mixture was refluxed for ten days. The mixture became ters possible by attack of the reagent upon a For example, Huender, Rcc. Ira% chim., 34, 26 (1916); Smith methyl group of the quinone. This mixture can- and(7)Homer, THIS JOURNAL, 62, 1349 (1940). not be separated into the pure components, but (8) Microanalyses by E. E. Hardy and E. E. Renfrew. JOURNAL, 62, 2825 (1940). (9) Smith and Guss, THIS it can be separated into two fractions, one of which (10) Smith and Harris, ibid., 67, 1289 (1935).

+

(5) Smith and Opie. THIS JOURNAL, 68, 937 (1941)

(11) Smith and Harris, ref. 10, p. 1292, give the m. p. as .56-0S0.

April, 1941

very dark, but no precipitate formed. Acidification followed by steam distillation gave tetraethylquinone (0.4 g., 80% recovery) and no other product could be isolated. B. Ethyl malonate ( 1 cc.) was added to a solution of sodium (0.156 g.) in dry ethanol (10 cc.). The quinone (0.5 9.) in dry ethanol (5 cc.) was added and the solution was warmed for several hours on the steam-bath. A succession of colors developed during the mixing of the reagents and the subsequent heating-fist red, then green, followed by brown and finally olive green, and a small amount of brown solid separated. This was removed, washed, with alcohol, then suspended in water and acidified. Only a very few flakes of solid material separated. The original alcoholic filtrate from this material was acidified and steam distilled. The quinone (m. p. 59-62', 0.36 g.; 72%) was the only product obtained although a small amount of uncrystallizable oil remained in the distilling flask. C. This experiment was repeated, using potassium instead of sodium. The results were the same; over 70% of the quinone was recovered and no other product could be isolated. D. Experiment A was repeated, substituting potassium for the sodium. Again the only product isolated was unchanged quinone. Trimethylethyl-$-benzoquinone (V).-5-Acetopseudocuniene12 was reduced to 5-ethylpseudocumene by the method of Smith and Kiess.l* The hydrocarbon (8.5 g.) was added 1 cc. a t a time to a vigorously stirred and well cooled (0') solution of potassium nitrate (20 g.) in sulfuric acid (280 cc.). The reaction mixture was poured over ice and the solid 3,6-dinitro-1,2,4,5-tetraethylbenzene(7.0 g., 50%) was removed and crystallized from methanol, when it melted a t 85-87'. Smith and Kiess14 obtained this nitro compound (m. p. 87-88') in 37% yield, but the method given here is decidedly superior to theirs. The use of a great excess of sulfuric acid and of potassium nitrate, and nitration as rapidly as possible, are essential for the success of the preparation. The dinitro compound was converted into trimethylethylquinone via the diamine stannichloride (which is appreciably soluble and must not be washed with large volumes of solvents) by the procedure of Smith and Kiess14 except that the stannichloride was not converted into the diamine but was oxidized directly to the quinone. The quinone was steam distilled from the reaction mixture, and isolated from the distillate by ether extraction. Removal of the ether left the quinone as yellow needles which melted a t 43-45O.16 Ethylpseudocumohydroquinone was prepared by boiling a solution of the quinone (2 g.) in acetic acid (10 cc.) and water (7 cc.) with zinc (20 mesh, 2 g.) until the mixture was colorless (thirty minutes), Boiling water (10 cc.) was added and the solution was a t once decanted from the zinc. The zinc was washed with hot water (10 cc.) and the liquid was decanted into the main solution. The hydroquinone separated from the cooled solution in practically quantitative yield. After crystallization from benzene, it melted a t 169-170'. A n d . Calcd. for CllH1602: C, 73.33; H, 8.88. Found: C, 73.58; H, 9.13. _____

~~~

(12) Smith (13) Smith (14) Smith (15) Smith

and and and and

935

TRIMETHYLETHYLQUINONE AND SODIUM MALONIC ESTER

Guss, THISJOURNAL, 69, 804 (1937). Kiess, ibid., 61, 284 (1939). Kiess, ibid., 61,993 (1939). Kiess, ref. 14, p. 994, give the m. p. a s 43'.

The diacetate, prepared in the usual way and crystallized from ligroin, melted a t 136-136.5'. Anal. Calcd. for C16H2004: C, 68.18; H, 7.57. Found: C, 68.38; H, 7.65. Trimethylethylquinone and the Sodium Enolate of Malonic Ester.-Ethyl malonate (19.2 g., 0.12 mole) was added to a suspension of powdered sodium (2.77 g., 0.12 mole) in dry, thiophene-free benzene (100 cc.) and the mixture was refluxed until all of the metal had reacted. The quinone (10 g., 0.06 mole) in benzene (50 cc.) was added and the mixture wasrefluxed. After a few minutes, a red color developed and after four hours a red solid began to deposit as a crust on the wall of the flask. The flask was shaken vigorously from time to time in order to break up this crust. Refluxing was continued for twentyfour hours, after which the mixture was cooled, the red sodium derivative was filtered and the adhering benzene was removed under reduced pressure. The red solid was triturated in a mortar successively with water, alcohol and ether. After drying, the solid weighed 10.4 g. (6O.8'%). The cooled benzene filtrates deposited white needles (1.13 g., equivalent to 11.16% of the starting material) of trimethylethylhydroquinone (m. p. 169-170"). After removing this, hydrochloric acid was added and the benzene solution was steam distilled. The benzene layer in the distillate was separated and the solvent, as well as the excess ethyl malonate, were removed under reduced pressure. There remained 1.75 g. of trimethylethylquinone (17,5070 of the starting material). Thus 89.46% of the quinone used was accounted for. 3 Carbethoxy 6 hydroxy bz dimethylethylcoumarin (VI -I- VI1 -I- VIII).-The red sodium derivative (3.7 g., 0.012 mole) was suspended in dry ethanol and hydrochloric acid (5 cc.) was added. The red color immediately gave way to the yellow color of the coumarin, and a white precipitate of sodium chloride formed, Water (100 cc.) was added and the coumarin ester (3.1 g., 90%) was removed and dried. I t melted a t 148-165'. Anal. Calcd. for C16HlsOs: C, 66.24; H, 6.20. Found: C 66.14, 66.52; H, 6.41, 5.97. The 185" Material (A).-I. From alcohol. The crude coumarin mixture (3.1 9.) was dissolved in alcohol (20 cc.) and the hot solution was filtered through a plug of glass wool. After standing for thirty minutes, the solution had cooled to 25' and a fluffy yellow precipitate (1.5 g., m. p. 177-180") had formed. This was removed and dissolved in 25 cc. of hot 1: 1 absolute ethanol and ligroin. The solution was cooled slowly to 25", the yellow solid (0.7 g., m. p. 185-186") was removed. The filtrate, when cooled to O", deposited a secord crop (0.35 8.) which melted a t 181-184". These two solids showed no depression in m. p. when mixed, hence the two were combined (total 1.05 9.) to form the product A. Anal. Calcd. for CleHlaOs: C, 66.24; H, 6.20. Found: C, 65.81, 65.68; H , 6.26, 6.32. Further concentration of any of the filtrates from this material led to deposition of lower melting solids, but these melted over a very wide range (153-160°, 150-165") and no improvement in the melting point value resulted when the solids were recrystallized repeatedly. 2. From ether-petroleum ether. The crude coumarin

-

-

-

-

-

936

LEE IRVINSMITH AND J. W. OPIE

ester was crystallized once by dissolving it in hot alcohol and cooling the solution t o 0'. The product melted over a wide range starting a t 147' and extending to above 170'. This material was shaken with ether, atered and petroleum ether was added t o the fltrate. The solution was concentrated until solid appeared in the hot solution, when enough ether was added to redissolve the solid. The resulting solution was allowed to cool slowly to 25' when it deposited material which melted at 178-180". When recrystallized again from ether-ligroin, this material melted a t 182.5-184'. It was possible by repeated ether extraction of the crude coumarin to obtain more of the substance melting at 184' but with every subsequent extraction the melting points of the extracted materials became progressively lower until finally only material melting at 140-150' was obtained. Further crystallization of these low melting products did not give any substances with higher melting points. 3. By chromatographic adsorption. Chromatographic adsorption on calcium carbonate or alumina (Brockmann or Baker) was of little value as a means of separating the coumarin isomers, although this method was very efficient in removing tarry impurities present in the crude coumarin mixture. On calcium carbonate, a brown layer a t the top of the column was held tenaciously, but practically all of the coumarin was unadsorbed. Evaporation of the filtrate gave material which melted at 172179'. When the coumarins were chromatographed from benzene onto Brockmann alumina, zones were formed as follows: a small brown zone, followed by a rather extensive pink and yellow zone and finally a pure yellow zone a t the bottom. Only the last gave any material on elution (alcohol), and from this there was obtained material melting a t 181-182' after one recrystallization from absolute ethanol. Coumarin Acid from A.-A solution of A (0.5 g., m. p. 179-180') in acetone (40 cc.) was refluxed for ninety minutes with hydrochloric acid (40 cc.). The coumarin acid (m. p. 214-222") was removed and crystallized from acetic acid. It then melted at 237-238". The 152' Material (B).-The coumarin mixture (obtained by decomposition of 10 g. of the red sodium derivative) was crystallized once by the ether-ligroin method, giving 3.53 g. of material melting a t 178-180'. The filtrates when set aside in a refrigerator for two months deposited material which melted a t 153-166'. This substance was washed with ether and the solutions, when evaporated, left material which melted at 150-152". The residue from the ether washing was dissolved in a larger quantity of ether and the solution was poured through a tube containing alumina. The filtrate was evaporated and the residue, after crystallization from ether-petroleum cther. melted a t 150-152".

VOl. ti3

Anal. Calcd. for ClaHlsOr: C, 66.24; H, 6.20. Found: C, 65.95; H, 6.24. Acetate.-The 152" material (0.5 9.) was dissolved in acetic anhydride (10 cc.). A drop of sulfuric acid was added and the solution was refluxed for fifteen minutes, then poured over ice. Excess ammonium hydroxide was added and the solid was removed and crystallized from dilute acetic acid (50%). The acetate melted a t 151152". When mixed with B (m. p. 150-152') it melted a t 126-130". Anal. Calcd. for C I R H ~ O O C,~65.04; : H , 6.02. Found: C, 64.88; H, 6.02. Coumarin Acid from B.-The 152' material B (0.5 g.) was dissolved in acetone (40 cc.), hydrochloric acid (40 cc.) was added and the mixture was refluxed for ninety minutes. The acid, removed from the chilled mixture and crystallized from dilute acetic acid, melted a t 230-232'. Anal. Calcd. for CIaH14Oi: C, 64.20; H, 5.43. Found: C, 63.93; H, 5.41.

Summary 1. Trimethylethylquinone has been condensed with sodium malonic ester. The initial product, a red sodium derivative, is formed in excellent yields. 2. The red sodium derivative, when decomposed by acids, gives a mixture of coumarin esters. This mixture was separated into two parts, A, melting a t 185O, and B, melting a t 150-152', by fractional solution and fractional crystallization. 3. Both A and B have been shown to be inseparable mixtures. B contains all three of the possible coumarin esters, VI, VI1 and VI11 derivable from trimethylethylquinone by reaction a t a methyl group, while A contains but two of the coumarin esters, VI and VII. 4. The ethyl group in trimethylethylquinone does not exert any pronounced orienting effect upon the reaction between the quinone and sodium malonic ester, such as is shown by the bromine atom in trimethylbromoquinone. 5. Tetraethylquinone did not react with sodium or potassium malonic esters under any of the conditions tried. MINNEAPOLIS, MINNESOTA RECEIVED DECEMBER 11, 1940